Background. Patient-delivered partner therapy (PDPT) is the practice of providing disease-specific antimicrobial agents to patients for delivery to their sex partners. Some partners who receive PDPT may forgo clinical evaluation, resulting in missed opportunities for the diagnosis and treatment of comorbid conditions.
Methods. We conducted a review of electronic charts for 8623 individuals attending 4 sexually transmitted disease clinics who were sex partners of patients with selected sexually transmitted infections (STIs). We evaluated the concordance between STIs diagnosed in partners and their reported exposures for which they might have received PDPT.
Results. Among 3503 female and 4647 heterosexual male partners, 19 (0.4%) of 4716 individuals tested were newly diagnosed with human immunodeficiency virus (HIV) infection, and 61 individuals (0.7%) had syphilis. Pelvic inflammatory disease was diagnosed in 133 women (3.8%). Seventy-two (3.2%) of 2226 female and heterosexual male partners reporting exposure to patients with chlamydial infection had gonorrhea diagnosed. Chlamydial infection or gonorrhea was diagnosed in 81 heterosexual male partners (10.3%) who reported contact with women with trichomoniasis. Among 473 men who have sex with men (MSM; including bisexual men), 13 (6.3%) of 207 tested were newly diagnosed with HIV infection, and 8 (1.7%) had syphilis. Six (6.1%) of 98 MSM reporting exposure to patients with chlamydial infection had gonorrhea diagnosed.
Conclusions. Infrequent coinfections in female and heterosexual male partners exposed to patients with chlamydial infection or gonorrhea would not preclude use of PDPT. However, PDPT for male partners of women with trichomoniasis and for MSM requires further study.
Partner notification is a long-standing component of public health and private-sector efforts to control sexually transmitted infections (STIs). Ideally, all sex partners of persons with an STI diagnosis would be notified of the exposure, evaluated, and treated. In practice, however, many patients with an STI diagnosis do not notify their sex partners of the exposure, and some partners do not access clinical services for STIs even if notified. Efforts to treat sex partners are further limited by how clinicians and health departments confront the issue of partner notification. Many clinicians outside public health settings do not consistently report cases of bacterial STIs to local health departments, and few notify their patients' partners directly [1–3]. Even when cases are reported, public health programs for partner notification provide services to <20% of persons reported to have Chlamydia trachomatis or Neisseria gonorrhoeae infection [4]. Consequently, many sex partners of persons with these infections remain untreated [5, 6], contributing to both ongoing transmission and high rates of reinfection [7]. As a result, there is widespread dissatisfaction with current approaches to partner management for STIs [8].
Recent studies have evaluated expedited approaches to partner management for chlamydial infection or gonorrhea that bypass the requirement for comprehensive clinical evaluation of partners. Such approaches include contracts between health departments and local pharmacies to provide medications for partners of patients with newly diagnosed STIs and to provide patient-delivered partner therapy (PDPT) [9, 10]. PDPT is the practice of providing disease-specific antimicrobial agents to patients for delivery to their sex partners, in addition to recommending referral for evaluation and care. Two observational studies found that PDPT was associated with reduced rates of persistent or recurrent chlamydial infection among women [11, 12]. More recently, 2 randomized, controlled trials compared strategies that included PDPT with those that primarily involved partner referral. The first of these found a nonsignificant reduction in persistent or recurrent chlamydial infection among women [10]. The second found significantly fewer persistent or recurrent chlamydial or gonococcal infections among men and women whose partners had received expedited therapy, including PDPT, compared with subjects whose partners had received only a referral to seek medical attention [13]. Although one study showed a reduction in recurrent trichomoniasis among women after treatment of their sex partners [14], data on the efficacy of PDPT in relation to trichomoniasis, syphilis, or other STIs have not been published.
Many private-sector clinicians [2, 15] and some public health departments [2, 16] use PDPT; however, the practice is not widely endorsed, and the 2002 Sexually Transmitted Diseases Treatment Guidelines published by the Centers for Disease Control and Prevention (CDC) [17] do not specifically recommend the use of PDPT. One barrier to greater acceptance of PDPT is the uncertainty about the potential for missed STI diagnoses among partners who forego clinical evaluation after they have been given medications. This study aims to define the potential missed opportunities to diagnose STIs among persons who might be candidates for PDPT and to assess the extent to which the impact of these missed opportunities might vary across the United States.
For the purpose of this article, use of the term “partner” will be limited to the description of our study subjects, who presented to sexually transmitted disease (STD) clinics as contacts of patients with a diagnosed STI. The term “patients” will describe individuals who were recently diagnosed with STIs; these individuals were not included as subjects within this study. We retrospectively analyzed data for individuals attending 1 of 4 public health STD clinics (in Baltimore, MD; Birmingham, AL; Denver, CO; and Seattle, WA) whose electronic medical records indicated that they were sex partners of patients with diagnosed chlamydial infection, gonorrhea, nongonococcal urethritis (NGU), mucopurulent cervicitis (MPC), or trichomoniasis. Partners in Seattle, Denver, and Birmingham were evaluated between 1 January 2001 and 31 December 2002. Partners in Baltimore were evaluated between 1 January 1999 and 31 December 2000. Individuals reporting exposure to patients with both chlamydial infection and gonorrhea were included; we excluded partners who reported exposure to patients with other combinations of STIs or HIV infection or syphilis and partners who could not identify the specific STI or syndrome to which they were exposed. For partners who were evaluated on multiple occasions during the study period, only data for the first episode were included. No effort was made to link data from partners with clinical information about their sexual contacts in the electronic medical record. All sites obtained either institutional review board (IRB) approval for research involving human subjects or IRB certification that the activity constituted program surveillance that was exempt from review.
Information on age, sex, race/ethnicity, sex of contacts, diagnosis received by contact, history of injection drug use, and number of sex partners was collected from the established databases or via abstraction from medical charts without identifying information. Male subjects were categorized as heterosexual if they reported contact only with women or as men who have sex with men (MSM) if they acknowledged sexual contact with any men. Male partners for whom the gender of their sex partners was not available in the database were excluded from the analysis.
Outcome variables—namely, diagnoses of HIV infection, syphilis, initial episode of genital herpes, and hepatitis virus infection, as well as pregnancy, contraception counseling, and Papanicolaou smear evaluation—were recorded from a combination of laboratory and clinical records. Pelvic inflammatory disease (PID) was diagnosed clinically, with each clinic using criteria specified by the CDC [16] or local modifications of the CDC criteria, without consideration of the results of laboratory testing or diagnostic imaging. A new diagnosis of HIV infection was based on a positive serological test result without a history of a prior positive test result. Subjects were classified as having early syphilis if either (1) the electronic medical record specified a clinical diagnosis of primary, secondary, or early latent syphilis or (2) they did not have a recorded history of syphilis, the nonspecific treponemal test titer was ⩾1: 16, and a Treponema pallidum-specific test result was positive. Subjects were classified as having late latent syphilis or syphilis of unknown duration if either (1) they had a clinical diagnosis of late syphilis or syphilis of unknown duration or (2) they did not have a recorded history of syphilis, the nonspecific treponemal titer was from 1: 1 to 1: 8, and a T. pallidum-specific test result was positive. Chlamydial infection or gonorrhea was diagnosed by either culture or nucleic acid amplification test. Trichomoniasis was diagnosed by microscopy at all sites and by culture at 1 site (Birmingham).
Some aspects of data collection and reporting varied across study sites. In Baltimore, men presenting with urethritis without evidence of gonococcal infection received a diagnosis of NGU without routinely being tested for chlamydial infection. Female partners who had been exposed to male patients with chlamydial infection were therefore classified as having been exposed to men with NGU. For this reason, data on STI diagnoses for partners presenting as having been exposed to patients with an STI in Baltimore were limited to PID, HIV infection, and syphilis, and data on chlamydial infection, gonorrhea, NGU, MPC, or trichomoniasis were not included. Of ∼5000 persons at the Birmingham clinic who had a history of relevant contact exposure, we included only 822 partners who were seen by a subset of 7 health care providers who used diagnostic codes for contact exposure consistently. Finally, the STD clinic in Seattle did not collect data on exposure of female partners to male patients with trichomoniasis; therefore, female partners in Seattle were not included in this subset analysis.
Concurrent STIs were evaluated in the context of antibiotic regimens previously studied for use in PDPT or otherwise considered to be standard for treatment of a specific STI. These single-dose regimens included azithromycin (1.0 g) for treatment of chlamydial infection, cefixime (400 mg) for treatment of gonorrhea, and metronidazole (2.0 g) for treatment of trichomoniasis. Treatment strategies for gonorrhea include treatment for concurrent chlamydial infection, whereas standard PDPT for partners of patients with chlamydial infection is limited to treatment for that STI.
Of ∼154,000 patient visits to the 4 clinics, 8623 individuals reported that they were sex partners of individuals with diagnosed chlamydial infection, gonorrhea, NGU, MPC, or trichomoniasis and met the other inclusion criteria. Study sites varied considerably in distribution by sex, sexual orientation, and race/ethnicity of partners and the STIs to which they were exposed (table 1).
Demographic characteristics of study subjects (partners) and prevalences of sexually transmitted infections (STIs) diagnosed in patients with whom they had contact.
Table 2 shows the prevalences of STI diagnoses among partners that were discordant with the diagnoses reported for the index patients. For example, partners who reported exposure to patients with chlamydial infection but who were found to have gonorrhea would not have received the standard treatment for their infection(s) through PDPT alone. Similarly, those partners who reported exposure to patients with trichomoniasis but who were found to have either chlamydial infection or gonorrhea would not have been adequately treated by PDPT for trichomoniasis.
Prevalences of pelvic inflammatory disease (PID), HIV infection, and sexually transmitted infection (STI) comorbidity among study subjects (partners) with a diagnosis discordant with the STI diagnosed in their contacts, by sex/sexual orientation of partners and clinic site.
Across all clinics, acute PID was diagnosed in 3.8% of female partners exposed to patients with chlamydial infection, gonorrhea, NGU, or trichomoniasis. HIV infection was newly diagnosed in 19 (0.4%) of 4716 female and heterosexual male partners who had been exposed to patients with STIs other than HIV infection or syphilis; however, all but 2 of these partners were from a single clinic (Baltimore). In contrast, across all clinics, 86 (10.3%) of 836 female and heterosexual male partners exposed to patients with trichomoniasis were found to have chlamydial infection or gonorrhea, and trichomoniasis was diagnosed in 99 (4.9%) of 2034 female partners exposed to patients with chlamydial infection, gonorrhea, or NGU. Among 473 MSM, 13 (6.3%) of 207 MSM tested were newly diagnosed with HIV infection. No significant differences in STI prevalences were found among MSM versus bisexual men (data not shown).
Across all clinics, infectious (early) syphilis was found in 28 (0.3%) of 8623 partners evaluated as having been exposed to patients with STIs. However, 24 of these 28 case patients were in Baltimore, where the prevalence was 0.7% (24 of 3225). The clinical diagnosis of early syphilis was made for 9 (32%) of the 28 subjects classified by the study as having early syphilis; the remaining 19 were classified on the basis of laboratory criteria.
Table 3 shows the concordance between STIs diagnosed in partners and the STIs to which they reported exposure. Combined data are shown for STD clinics in Birmingham, Denver, and Seattle. Chlamydial infection was diagnosed in 1025 (44.1%) of 2324 partners of patients with chlamydial infection, and prevalence ranged from 4 (12%) of 32 MSM in Denver to 567 (53%) of 1077 heterosexual men in Denver (data for individual clinics not shown). Gonorrhea was diagnosed in 400 (38.8%) of 1032 partners of patients with gonorrhea, and prevalence ranged from 7 (13%) of 55 heterosexual men in Birmingham to 141 (57%) of 248 women in Denver. Chlamydial infection was diagnosed in 173 (16.8%) of 1032 partners of patients treated for gonorrhea, and prevalence ranged from 7 (5%) of 135 MSM in Denver to 66 (27%) of 248 women in Denver. The prevalence of gonorrhea among partners of patients treated for chlamydial infection was lower and varied from 2 (1%) of 147 women in Seattle to 4 (12%) of 32 MSM in Denver.
Prevalences of pelvic inflammatory disease (PID), HIV infection, and sexually transmitted infections (STIs) diagnosed in study subjects (partners), by STI diagnosed in contacts.
Forty (0.5%) of 8623 partners had a diagnosis of an initial episode of genital herpes, and 18 (0.4%) of 4576 tested positive for hepatitis C virus infection at the 2 clinics where testing was offered. Papanicolaou smears were collected from 464 (22.2%) of 2088 women at 3 clinics, and results for 109 of the smears were abnormal. Hormonal contraception was given to 148 (8.6%) of 1726 women, and 115 (5.9%) of 1948 women had a positive pregnancy test result. Hepatitis vaccination was administered to 68 (3.2%) of 2109 subjects.
Many private-sector providers and a few public health departments nationwide use PDPT at least intermittently [2]. Our study identified relatively low rates of STIs that might not be diagnosed and treated through use of PDPT among female and heterosexual male partners of patients with chlamydial infection or gonorrhea, although ∼1% of partners in Baltimore had newly diagnosed HIV infection. This prevalence was lower than the prevalence of HIV infection among all persons tested at the STD clinic in Baltimore during the study period, but it may be high enough to warrant caution in using PDPT in settings with relatively high rates of heterosexual HIV transmission. Additional research is needed to identify patient characteristics that may be predictive of a high risk for HIV infection or other major comorbidities in partners who might be potential candidates to receive PDPT.
A small percentage of partners of patients with chlamydial infection had a clinical diagnosis of PID. The diagnosis of PID would typically prompt additional therapy that would not be included as part of PDPT for exposure to chlamydial infection or gonorrhea, and the possibility of inadequate therapy for PID could be considered a potential cost of PDPT for female partners exposed to men with chlamydial infection or gonorrhea. We therefore would recommend that both written and verbal instructions be delivered to female sex partners, advising them of the symptoms of PID and the need to seek medical evaluation if they have those symptoms.
Similarly, 3.8% of female partners and 3.0% of heterosexual male partners of patients with chlamydial infection were found to have gonorrhea. Most cases of gonorrhea would be adequately treated with a 1.0-g dose of azithromycin given as treatment for exposure to chlamydial infection [18, 19]. However, some clinicians and health departments provide doxycycline, instead of azithromycin, as treatment for chlamydial infection because of cost. Although many cases of gonorrhea would also respond to a course of doxycycline, ∼15% of gonococcal isolates in the United States are resistant to doxycycline [20]. In addition, compliance with PDPT can be maximized by the use of single-dose regimens, and PDPT with doxycycline would require 1 week of therapy. In the absence of prospective studies of the use of doxycycline for PDPT for chlamydial infection, we believe that azithromycin should be regarded as the first choice for PDPT for chlamydial infection.
Despite the lack of evidence for the efficacy of PDPT with metronidazole for male partners of women with trichomoniasis, trichomoniasis is the most frequent STI for which PDPT has been used. In this context, the finding that chlamydial infection or gonorrhea occurred in 81 (10.3%) of 785 male partners of women with trichomoniasis was a concern. This finding may have resulted, in part, from misclassification due to error in the partner's reported exposure. Some women with trichomoniasis who referred partners for evaluation may have had multiple STIs but would have been aware only of the trichomoniasis diagnosis at the time that they notified their partner(s). Partners exposed to patients with both trichomoniasis and a bacterial STI therefore might have been informed only of the exposure to trichomoniasis and might have reported contact with only that infection at the time of their evaluation. Thus, the prevalence of chlamydial infection or gonorrhea among male partners exposed to women with trichomoniasis warrants further study in different settings. At a minimum, women also should be tested for chlamydial infection and gonorrhea before clinicians provide PDPT for trichomoniasis for male partners.
Conversely, public health STD clinics serving patient populations with a relatively high prevalence of trichomoniasis should consider the potentially lost opportunity for detection and treatment of trichomoniasis as one cost of instituting a PDPT program for chlamydial infection or gonorrhea. However, many patients presenting to public health STD clinics or to other clinicians after exposure to other STIs probably are not tested for trichomoniasis, and no national program currently exists to control this infection. Therefore, although missed opportunities to diagnose trichomoniasis are a potential cost of the use of PDPT for partners exposed to chlamydial infection or gonorrhea, we do not believe that this cost necessarily outweighs the benefit of the practice.
Our study raises concerns about the use of PDPT among MSM. Of 473 MSM in our study, 46 (9.7%) were previously known to be HIV positive, 207 (43.8%) were tested for HIV infection, and 13 (6.3%) of these 207 were newly diagnosed with HIV infection. An additional 0.4% had early syphilis, and 1.6% had late latent syphilis or syphilis of unknown duration. In contrast to infection in women, chlamydial infection and gonorrhea are seldom associated with long-term sequelae in men. The use of PDPT for MSM might decrease the community burden of STI and perhaps have a secondary effect on individual- and community-level risks of HIV acquisition and transmission. The magnitude of these potential benefits, if any, is unknown. Given the significant consequences of missed diagnoses of HIV infection and syphilis and the absence of any data on clinical outcomes, we believe that PDPT for chlamydial infection, gonorrhea, or NGU should be used with caution with MSM until further studies have been performed.
Several factors may limit the generalizability of our findings. Although we selected study sites that represent diverse regions of the United States, including areas with substantial HIV epidemics among heterosexual individuals, levels of STI comorbidities may be higher in some settings. In addition, all the subjects in our study sought medical care in STD clinics, whereas not all partners informed of their exposure to an STI will seek care, and many will not seek care at public health STD clinics.
In conclusion, in acknowledging that some partners will never seek care and that many public and private clinics lack resources to provide comprehensive partner-notification assistance, PDPT may increase the proportion of partners treated for exposure to STIs. The decision to use PDPT requires clinicians to weigh the proven benefit of PDPT against the risk of missed opportunities to counsel partners and to identify and treat comorbidities. In our study, STI comorbidities were uncommon among women and heterosexual men who sought treatment after exposure to an individual with chlamydial infection or gonorrhea, but comorbidities were more common among male partners exposed to women with trichomoniasis and among MSM. Although undiagnosed HIV infection and early syphilis in patients evaluated at the STD clinic in Baltimore were uncommon, relative to the prevalences found among MSM and bisexual men, the costs and benefits of PDPT for heterosexual partner pairs in this and similar settings may warrant further operational research, including evaluation of possible strategies to help motivate further care-seeking behavior.
We thank Fred Koch and Shuying Yu for their help with data management.
Financial support. National Institutes of Health (grants T32 AI07140-26 [to J.S.] and K23 AI01846 [to M.R.G.]).
Potential conflicts of interest. All authors: no conflicts.
IDSA Members: For your free access to this journal, log in via the IDSA members area.
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy